The Impact of 2000 Census-based Population Controls on Health Estimates in the National Health Interview Survey

• Activity Number: 21
• Type: Contributed
• Date/Time: Sunday, August 8, 2004 : 2:00 PM to 3:50 PM
• Sponsor: Section on Health Policy Statistics
• Author(s): Carrie Lynch*+ and Van L. Parsons
• Companies: National Center for Health Statistics and National Center for Health Statistics
• Keywords: multiple-race ; population controls ; weights ; standard errors

Abstract

In 1997 the Office of Management and Budget (OMB) issued revised standards for the collection and presentation of federal data on race and ethnicity. The 2000 Census adopted these standards, and other federal surveys were to comply by January 2003. The National Health Interview Survey (NHIS), which is a major national household survey that monitors the health of the U.S. civilian noninstitutionalized population, has been allowing respondents to select more than one race since 1976. The NHIS also adopted the revised OMB standards with the fielding of the 2000 NHIS, but until 2003, the survey weights were produced using 1990 Census-based population controls and old race categories. This paper describes the procedure of bridging 2000 Census based population controls distributed by multiple-race groups to single-race groups for the development of NHIS weights. This paper also compares selected health estimates using 1990-based weights and 2000-based weights. The results presented in this paper demonstrate that, in practice, the number of significant differences of proportions is negligible. However, the switch to new population controls affects total estimates.